We propose a methodology to create lines of mice that are capable of the robust expression of any transgene in an anatomically specific manner. We call this method "subtractive transgenics", because it involves (by a combination of two proven transgenic technologies) the subtraction of the anatomical specificity of one expression pattern from that of another. We plan to use these mice to dissect out the functional circuitry of the central nervous system (CNS) by expressing "silencers" (constructs which turn neurons off) with unprecedented anatomical specificity. This should allow us and others to analyze the neural circuits of the CNS in a way analagous to how an engineer analyzes an electronic circuit: short out one element, and then record what happens to downstream elements. The first few specific constructs we propose should result in transgene expression in various parts of the forebrain that have been implicated in neuropathologies as diverse as Alzheimer's disease and other dementiae (the hippocampal formation) to Parkinson's disease and other disorders of the basal ganglia (striatum). The heightened understanding of the functional circuitry of the CNS the resulting mice will enable will lead to a better understanding of the etiology of its pathological states, and allow for the generation of better mouse models of these human disorders. However, it should be stressed that the method can increase the anatomical specificity of transgene expression in any tissue type, and should therefore be useful to biomedical research in general. PUBLIC HEALTH RELEVANCE. The mammalian brain is the most anatomically-complex structure in nature (and we have the most complex mammalian brain), composed of innumerable electrical interactions between literally thousands of different cell types. It is an incredibly complex biological circuit, in essence, and the cell types are its component parts. Many neurological and psychiatric disorders can be thought of as imbalances in different parts of this central circuitry. We propose a method to create genetically-modified lines of mice that can express transgenes in different specific areas of these central circuits, to enable researchers to try to understand what the different parts do. The transgenes that can be expressed range from things that just turn the cells off, as we propose to do, to specific genetic disease models. The first few specific constructs we propose should result in transgene expression in various parts of the forebrain that have been implicated in neuropathologies as diverse as as Alzheimer's disease and other dementiae (the hippocampal formation) to Parkinson's disease and other disorders of the basal ganglia (striatum). The heightened understanding of the functional circuitry of the mammalian brain the resulting mice will enable will lead to a better understanding of its pathological states, and allow for the generation of better mouse models of human disorders.